inkspot: Two-way ordered bubble plot of a species by sites data table

In nsj3/rioja: Analysis of Quaternary Science Data

Two-way ordered bubble plot of a species by sites data table

Description

Plots a two-way ordered bubble plot of a species by sites data table, where species are rows and sites are columns. The sites can be ordered and the functions will sort species to cluster abundances on the diagonal.

Usage

inkspot(data, gradient=1:nrow(data), use.rank=FALSE, 
      reorder.species = TRUE, x.axis=c("sites", "gradient", 
      "none"), x.axis.top=FALSE, site.names=NULL, spec.names=NULL, 
      pch=21, cex.max=3, col="black", bg="darkgrey", 
      x.grid=FALSE, y.grid=FALSE, grid.col="grey", grid.lty="dotted",
      legend.values=c(2, 5, 10, 20, 50), ...)

Arguments

data	data frame to be plotted.
gradient	a vector for ordering sites along the x-axis.
use.rank	logical to indicate that the rank rather than absolute values of gradient should be used to plot site positions. Defaults to FALSE.
reorder.species	should species be reordered to reflect pattern in site ordering? Defaults to TRUE.
x.axis	controls labelling of x.axis. sites uses site names, gradient plots an axis relecting values of the supplied gradient, and none omits labels and draws ticks at the site positions.
x.axis.top	logical to include an x-axis on the top of the figure labelled with values of the gradient.
site.names, spec.names	character vectors of site or species names to annotate the axes. Defaults to row and column names.
cex.max	maximum size of plotting symbol. Symbols are scaled so maximum species abundance has a symbol of this size.
pch, col, bg	plotting symbol and line / fill colour.
...	additional arguments to plot.
legend.values	if not null, places a legend in the top-left corner displaying the listed values.
x.grid, y.grid	draw grid at x and y ticks.
grid.col, grid.lty	grid colour and line type.

Details

Function inkspot plots a two-way table of species by sites as a bubble plot, with sites ordered along the x-axis, species on the y-axis, and abundance indiacted by scaled symbols ("bubbles"). It is a useful way to visualise species distribution along an envionmental, spatial or temporal gradient. If gradient is not given sites are plotting in the order they appear in the input data. Otherwise sites are plotting according to the values in gradient. If site labels overlap (multiple sites at similar values of gradient), labels can be suppressed x.axis= "none", or replaced with the gradient axis x.axis="gradient". A gradient axis can also be added to the top x.axis (x.axis.top=TRUE. Symbols are scaled so that the maximu abundance has a symbol size of cex.max. If sites are spaced unvenly along the gradient, or if many labels overlap, sites may be plotted evenly spaced using use.rank=TRUE. In this case the function will place top axis labels (if requested) at the appropriate positions along the gradient.

Value

Function inkspot returns a list with two named elements:

spec	index of the species order.
site	index of the site order.

Author(s)

Steve Juggins

See Also

vegemite in package vegan for a tabular alternative.

Examples

data(SWAP)
mx <- apply(SWAP$spec, 2, max)
spec <- SWAP$spec[, mx > 10]
#basic plot of data with legend
inkspot(spec, cex.axis=0.6)

#order sites by pH
pH <- SWAP$pH
inkspot(spec, pH, cex.axis=0.6)

# add a top axis
inkspot(spec, pH, x.axis.top=TRUE, cex.axis=0.6)

# order by pH but plot sites at regular intervals to avoid label overlap
inkspot(spec, pH, use.rank=TRUE, x.axis.top=TRUE, cex.axis=0.6)

# or add long taxon names
oldmar <- par("mar")
par(mar=c(3,12,2,1))
nms <- SWAP$names[mx > 10, 2]
inkspot(spec, pH, spec.names=as.character(nms), use.rank=TRUE, 
x.axis.top=TRUE, cex.axis=0.6)
par(mar=oldmar)

nsj3/rioja documentation built on Sept. 17, 2024, 5:50 a.m.